__jochen__

nice, but did you expect it to not work on Safari (WASM fails)?

Sorry, I'm going to be a downer. I've been around for a long time too, saw many formats come and go (even contributed to some myself). I think Markdown is super neat and handy, but this statement "The trillion-dollar AI industry's system for controlling their most advanced platforms is a plain text format... Their achievement is every bit as impressive as yours." ..is way off. NN math & engineering has been refined for ~50 years (give or take) and scaled to mindboggling levels. For better or worse, it is in the process of transforming how society functions (just like the internet and mobile phones did). Building modern advanced NN/AI requires extremely sophisticated and advanced science, hardware & algorithms; the format of the prompts conventionally used by some are a handy but fairly trivial part of the endeavor.

From a technical perspective there is so much you can do to secure things, including against theft. While it's hard to lock things down past 99.9% usage, it is not too hard to make token issuance secure enough for practical, wide-spread use (there are a plethora of crypto protocols out there to prove the point).

There's no guarantee that the government will pick the best standard, but one can hold out hope (e.g. when the US govt adopted Rijndael as the AES encryption standard).

Amazing! Do you have some spare time? Can you quickly mint me a BTC token please. kthxbai ;)

The proposal is for SIGNED tokens i.e. only the govt can issue them, and you need a govt issued ID to generate them. The latter mechanism allows rate limiting. This fixes the problem you outline.

That's the problem. The new norm will be 10x of pre-AI productivity, nobody will be able justify hand-writing code. And until the quality bar of LLM's/their successors get much better (see e.g. comments above looking at the details in the examples given), you'll get accumulation of errors that are higher than what decent programmers get. With higher LOC and more uninspected complexity, you'll get significantly lower quality overall. The coming wave of AI-coded bugs will be fun for all. GOTO FAIL;

There is also likely to be increased pressure in a SE job to produce more code. You'll find that if others use AI, it'll be hard to be a hold-out and hit fewer delivery milestones, and quality is hard to measure. People are rewarded for shipping, primarily (unless you're explicitly working on high reliability/assurance products).

Agreed, that this is a limited scope for what AF does. Nevertheless, it seems interesting to confirm that its primary strengths are memorization, which is not surprising.